In this session, you will learn about what you should do after you’ve taken an AI transformation baseline. Over the span of this session, we will discuss the next steps in moving toward AI readiness through alignment of talent and tools to drive successful adoption and continuous use within an organization.
To find additional videos on AI courses, earn badges, join the courses at H2O.ai Learning Center: https://training.h2o.ai/products/ai-foundations-course
To find the Youtube video about this presentation: https://youtu.be/K1Cl3x3rd8g
Speaker:
Chemere Davis (H2O.ai - Senior Data Scientist Training Specialist)
AWS Community Day CPH - Three problems of Terraform
AI Foundations Course Module 1 - Shifting to the Next Step in Your AI Transformation Journey
1. AI Foundations
Module 1: An AI Transformation Journey
Session 2: Shifting to the Next Step in the AI Transformation Journey
2. 2
What you
can expect
in this
session
01 Introduction
02 Four Phases of AI Maturity
03 Success With AI
04 What’s Next?
3. 3
Module 1: An AI Transformation Journey
Session 1: An AI Transformation Journey Primer (replay posted)
Session 2: Shifting to the Next Step in the AI Transformation Journey
Study Group
Ask Me Anything Forum
Session 3: AI Transformation and Covid-19 (Special Session)
Module 2: Demystifying AI
Module 3: Machine Learning Foundations
AI Foundations Overview
Interested in knowing the full
schedule for the AI Foundations
course? View the schedule on
the community learning site
You Are Here
6. 6
Who is on the team?
Business leader, data
scientists, IT professional
Determine the problems you
want to solve with metrics (time,
money, # of customers, etc)
Determine where you have
data, need data, and can
use technology to find
answers and predictions.
Find answers efficiently.
Learn from others in the
data science community
Ask the Right Questions
Data & Technology
Community
Create a Data Culture
Understand and explain the
models. Use leading edge
technologies to guard for
bias, explain a model, and
present this to regulators
Trust in AI
2
1
3
4
5
5 Keys
to unlock AI
7. 7
AI Business Value a Journey in Four Phases
1 2 3 4Potential Operational Strategic Data-Driven
Enterprise AI Journey
Awareness & Interest
Evaluate Business Value
Technical Evaluation
Point
Deployment
Point
Production
Enterprise Deployment
Enterprise Production
Modern AI Architecture
Industry Leadership
8. Confidential8 Confidential8
Engagement Cycle (An Example)
d
1. Data Assessment 2.Use Case Workshops
With AI Experts
3.Accelerated Model
Development via Automatic
Machine Learning
4. Model Deployment
5. Measurement of KPI
(Revenue Increase or Cost
Decrease)
Team should rapidly iterates through 5 key steps alongside stakeholders for each use case to focus on
tangible, measurable results and turnkey work products
10. Confidential10
AI Journey Example
Value
Time
Partner
Initial Use Case
Drives Value
AI Driven
Business
0
(Baseline)
Digital Transform &
Data Management
Initiatives Begin Basic Value From
Organized/Structured Data
Begins to Break Even
Rapid Use Case
Development Phase
Multiple Use
Cases Move
into
Production
11. Confidential11
AI Business Value a Journey in Four Phases - Roles
1 2 3 4Potential Operational Strategic Data-Driven
AppsInfrastructureBusiness
CXO, CDO, CIO, Line of Business Owner
Business Analyst, Developers, Data Scientists, Data
Workers, DBAs
Data Architect, Dir. Of Security, Platform Operations, devOps, DBAs
12. Phase 1: Potential
Module 1: An AI Transformation Journey
Session 2: Shifting to the Next Step in the AI Journey
Part: 2
13. Confidential13
Where are You? Maturity at Each Phase of the Journey
1 2 3 4Potential Operational Strategic Data-Driven
AppsInfrastructureBusiness
Focus on Entry
Application
Validate the
Value of AI
14. Confidential14
Where are You? Maturity at Each Phase of the Journey
1 2 3 4Potential Operational Strategic Data-Driven
AppsInfrastructureBusiness
Data Scientist
Data Engineer
15. Phase 2: Operational
Module 1: An AI Transformation Journey
Session 2: Shifting to the Next Step in the AI Journey
Part: 3
16. Confidential16
Where are You? Maturity at Each Phase of the Journey
1 2 3 4Potential Operational Strategic Data-Driven
AppsInfrastructureBusiness
Focus on Entry
Application
Validate the
Value of AI
Deploy your next
set of App(s)
Choose Initial AI
Advance
Analytics App(s)
17. Confidential17
Where are You? Maturity at Each Phase of the Journey
1 2 3 4Potential Operational Strategic Data-Driven
AppsInfrastructureBusiness
Data Scientist
Data Engineer
Model Ops
App
Developer
Business
Decision Maker
Project1
Data Scientist
Data Engineer
Model Ops
ProjectN
Business
Decision Maker
Module: 6
18. Phase 3: Strategic
Module 1: An AI Transformation Journey
Session 2: Shifting to the Next Step in the AI Journey
Part: 4
19. Confidential19
Where are You? Maturity at Each Phase of the Journey
1 2 3 4Potential Operational Strategic Data-Driven
AppsInfrastructureBusiness
Focus on Entry
Application
Validate the
Value of AI
Deploy your next
set of App(s)
Choose Initial AI
Advance
Analytics App(s)
Executive
Sponsorship
for organizational
transformation
Deploying
Multi-tenant AI
Platform
Centralize AI
Data Governance
& Security
21. Confidential21
Rich AI Ecosystem - Too Many Choices?
Databases Big Data/Distributed
Computing
Cloud Computing
Programming Languages Business Intelligence Data Science/Analytics/AI
• Typical frontline store
of data (relational,
graph, etc)
• May be hosted in
cloud if volume of data
warrants it
• If data is too big to be
useful for accessing it
you can use big data
platforms for
distributed, parallel,
high-performance
computing
• In terms of accessing,
isolating, cleaning,
transforming data,
these are the big 3.
• Python + R are
consistently used for
DS & modeling
• Most common
resources for
descriptive statistics
and dashboarding
(specialize in
descriptive stats)
• For predictive & advanced analytic
insights use Data Science/AI
platforms (and py+R) to apply the
highest quality methods.
• Cloud computing may be needed
to run heavy math for these
models.
Module: 5
ML
Foundations
ML
Foundations
22. Confidential22
Where are You? Maturity at Each Phase of the Journey
1 2 3 4Potential Operational Strategic Data-Driven
AppsInfrastructureBusiness
Data
Architect
Business
Unit1
Business Analysts
Application Dev
Lead
Model Ops
Data
Engineer
Enterprise
Data Science
Team
BI
Teams
Divisional
Governance
Dir, IT Ops
Application Dev
Lead
P1
P2
PN
Business
UnitN
Business Analysts
Application Dev
Lead
Model Ops
Data
Engineer
Enterprise
Data Science
Team
BI
Teams
Divisional
Governance
Dir, IT Ops
Application Dev
Lead
P1
P2
PN
23. Confidential23 Confidential23
Two Sides of the Same Coin
Understanding the entire
scope of risks of deploying
AI is imperative for success
Risk
For firms to fully adopt &
take of AI, trust & buy-in
critical
Trust
Correctly understanding & quantifying risk increases trust
Knowing how & when to trust your AI decreases risk
Business leaders are focused on decreasing risk, while data scientists are focused on understanding and
trust of these models. Ultimately, these goals go hand in hand in hand.
24. Confidential24
Responsible AI Responsible AI
Ethical AI
Secure AI
Explainable
AI
Interpretable
Machine
Learning
Human-Centered
Machine
Learning
Compliance
RESPONSIBLE
AI
As the field has evolved, many definitions and concepts have come
into the mainstream, below we outline H2O.ai’s respective definitions
& understanding around the factors that make up Responsible
Artificial Intelligence.
• Explainable AI: Focuses on the ability analyze a ML model after it has been
developed
• Interpretable Machine Learning: Transparent model architectures and increasing
how intuitive and understandable ML models can be
• Ethical AI: Sociological fairness in machine learning predictions (i.e., whether one
category of person is being weighted unequally)
• Secure AI: Debugging and deploying ML models with similar counter-measures
against insider and cyber threats as would be seen in traditional software
• Human-Centered ML: User interactions with AI and ML systems
25. Confidential25 Confidential25
Data Loading, Prep,
Exploratory Analysis,
and Feature
Engineering
Model Training &
Assessment
Review for
Explainability,
Interpretability, and
Fairness
Deploy Model in
Production
Start Pass Pass Pass
Responsible AI Process Overview
Published by H2O.ai
Ongoing Review and Monitoring
Steps where data/statistical/human bias
can enter the process
Fail & Review
26. Phase 4: Data-Driven
Module 1: An AI Transformation Journey
Session 2: Shifting to the Next Step in the AI Journey
Part: 5
27. Confidential27
Where are You? Maturity at Each Phase of the Journey
1 2 3 4Potential Operational Strategic Data-Driven
AppsInfrastructureBusiness
Focus on Entry
Application
Validate the
Value of AI
Deploy your next
set of App(s)
Choose Initial AI
Advance
Analytics App(s)
Executive
Sponsorship
for organizational
transformation
Deploying
Multi-tenant AI
Platform
Centralize AI
Data Governance
& Security
Insights from
Data becomes a
Competitive
Advantage
Widespread
Adoption &
Consumption
Continuously
Refine AI Best
Practices
28. Confidential28
Where are You? Maturity at Each Phase of the Journey
1 2 3 4Potential Operational Strategic Data-Driven
AppsInfrastructureBusiness
AI
Architect
Model Ops
Data
Workers
Data Science
Team
Corporate
Governance
Dir, IT Ops
Application Dev
Lead
P1 P2
CDO
BU1
BU1
Business
Analysts
BI
Teams
Business
Analysts
BI
Teams
P1 P2
P1 P2P1 P2
CTO
Application
Architect
End User Service
Desk
Infrastructure
Administration
Change & Program
Mgt.
29. Confidential29
Where are You? Maturity at Each Phase of the Journey
1 2 3 4Potential Operational Strategic Data-Driven
Data or Capital?
Market Capitalization
Peer Group
Leaders
A Mature Data Driven Organization treats Data like Capital
Chief Financial Officer
Managed for shareholder value
Top-down mandate to allocate capital
Cost control pushed to front-line
Cost of capital
Every dollar can be invested once
Value compounds over time
Chief Data Officer
Managed for shareholder value
Top-down mandate for a data-driven organization
Data-driven decision making pushed to
front line
Cost of data processing infrastructure and
expertise
Data is an organizational asset
30. Organize for Success With AI
Module 1: An AI Transformation Journey
Session 2: Shifting to the Next Step in the AI Journey
Part: 6
31. Confidential31
Organizing your AI Journey for Success
Enable self-service for the business
• Analysts, data scientist and developers aligned with the particular needs of each business
• AI consumable for different use cases/workloads
• Multiple lenses on the same data, with team specific views
Central operational constructs
• Better ability to attract, retain and develop specialist staff
• Consistent data science best practice enforcement and compliance
• Optimal capital efficiency driven by scale and a multi-tenant shared service
• Modular, enterprise architecture that supports the broadest range of applications and analyses
32. Confidential32
What We’ve Covered So Far...
• AI Transformation requires a journey through the phases of maturity:
– Potential
– Operational
– Strategic
– Data-Driven
• Embracing the 5 Keys to Unlocking AI play a crucial role in moving to maturity
• As maturity increases, data becomes an organizational asset that can be turned into real business
value
• AI adoption improves when there is clear strategy and alignment with talent, technology, and goals.
Recording will be
posted w/in 2
days
33. What’s Next
Module 1: An AI Transformation Journey
Session 2: Shifting to the Next Step in the AI Journey
Part: 7
34. Confidential34
1. Study Group on Saturday July 4, 2020 @ 7:00AM PDT
2. Ask Me Anything on Sunday July 5, 2020 @ 7:00AM PDT
3. The next session AI Transformation & Covid-19 will be held on
Monday July 6, 2020 @ 7:00AM PDT
4. Module 2: Demystifying AI will begin on Tuesday July 7, 2020 @
7:00AM PDT.
Upcoming Sessions
35. Confidential35 Confidential35
Quizzes & Study Groups
• Each session within a module will have a small quiz to complete and all
quizzes for that module will be due before the next module starts.
• There are 2 options available for you to ask additional questions or get
assistance on AI concepts covered in the sessions:
– A Study Group for each Module will be held on Saturdays @ 7:00AM PDT
– Ask Me Anything will be held on Sundays @7:00AM PDT
• Reminder: Don’t forget to complete Quiz 2: Shifting to the Next Step in
the AI Transformation Journey by Tuesday July 7, 2020 to earn your
badge!
37. Confidential37 Confidential37
Additional Resources
H2O.ai’s AI Glossary
H2O.ai Webinar - Maximizing the Impact of Machine Learning through Model Ops
H2O.ai Webinar - Key Terms and Ideas in Responsible AI
H2O.ai Webinar - Your AI Transformation